According to the cooperative principle everything one says in a conversation has to be not only relevant but also informative, truthful and perspicuous. This is entailed by the 'supermaxims' of 'relation' (that is, relevancy), 'quantity', 'quality' and 'manner'. It has been suggested that the supermaxims of quantity, quality and manner are simply 'subordinated to the general maxim of relevance' and that the whole general principle of conversation is 'in fact a principle of relevance rather than a principle of cooperation'. This suggestion is certainly correct for the supermaxim of quantity which reads that one should make one's contribution as informative as is required, and not more informative than is required: this is just informational relevance. But the suggestion misses the point of the supermaxim of quality which reads that one should make one's contribution one which is true.

On the surface it is understandable that someone would argue that a relevant conversational contribution must also be a true one, and that an untrue contribution cannot be relevant. Yet, this reasoning is erroneous, since the entity which is supposed to be relevant is not the same one as the entity which is supposed to be true. Consider, by way of an example, someone who says "it's raining" and assume that this proposition or statement is true at the place and time 'e utters it. It is then true that it is raining, and false that it is not raining. Now, assume also that the person's contribution to the conversation may be called "relevant" at the same place and time. The 'relevance' of the contribution or statement or proposition is then only a derivative one, because the statement or proposition derives its relevance from its content in the given context. Is it then the fact that it is raining itself which is relevant? No, because if it is, or were, relevant that it rains or rained, it is, or would be, relevant too that it does or did not rain. What is relevant is the question whether it is raining or not. In other words, the distinction between raining and not raining. A statement making use of this distinction is relevant at the time and place concerned so far as this distinction is concerned. If the speaker had said "it's not raining", this would have been relevant as well, but then it would have been false and a violation of the principle of truth (or of the 'supermaxim of quality', if preferred). Consequently, this principle can definitely not be dispensed with, even by those who pay the fullest attention to questions of relevancy.

It has also been argued by certain philosophers of language that the cooperative principle for conversations is a 'maxim of simple relevance which would constrain the speaker's choice of utterance, and the hearer's choice of interpretation, hardly at all'. They would substitute a maxim of maximal relevance for the cooperative principle. What they are concerned with is not so much one relevancy relation which plays a role in one context in a narrow sense (to be compared with the phenomenologist's domain of relevance) but rather with a multiplicity of relations to different goals or topics (to be compared with the phenomenologist's system of relevance). (See for the schematic representation of this view figure I.5.2.3.1 again.) The underlying idea is that a conversation or discourse involves a sequence of usually many propositions, and never one proposition (or utterance) in isolation. This idea does not contradict the conception of 'simple' relevancy. On the contrary, it builds on it and extends it.

So far as discriminational relevancy is concerned it is not our immediate concern whether people work in practise with notions of simple relevancy or with notions of more complex relevancies. What is far more interesting from this point of view is that the type of relevancy (in any way pragmatic relevancy) in questions of conversation and information appears to be discriminational as well. It is again a distinction which is relevant or not, and it is the presence of this distinction in a statement or proposition which can make this statement or proposition itself relevant. It is not only the presence of this distinction in a whole sentence, it starts with its presence or absence in each separate morpheme (such as the presence of gender in brother and sister and the absence of it in sib and sibling). Whether the ultimate relevance lies in the distinction or in the factor is not important, but given the relevance of the distinction or factor, the relevance of a morpheme, of a word, of a proposition and of uttering a proposition derives from it. The same holds for the so-called 'making of a distinction': making a distinction is relevant because the distinction is relevant, and not the other way around. Whether a distinction is drawn or not is not a matter of relevancy but of truth. Granted that a distinction is relevant, drawing it is relevant, and granted that it is not relevant, drawing it is not relevant.

Pragmatic relevancy in particular is thus nothing else than a form of discriminational relevancy in the conversational or informational field. What we have been calling "discriminational relevancy" hitherto was basically the discriminational relevancy of the ground-world in which nonpropositional distinctions exist, and are made, between nonpropositional entities or classes of such entities. Unlike this form of discriminational relevancy, the role played by relevancy in questions of conversation and information can solely be understood against the background of propositional reality. That is why it is a subject of philosophy of language, of logics, of philosophy of science and of other disciplines concerned with communication, thought and valid theorizing. So it has been noticed that it is not 'of the tradition of science nor of its spirit to give irrelevant information'. Scientific information, it is said, must be relevant to the topic, and science is not just a question of citing all the knowledge one has (even tho the suggestion is part of the etymological origin of the word science). But if all information and conversation has to be both relevant and true, it violates the principle of relevance itself to exclusively emphasize that scientific information and conversation has to be relevant and true. If it did not violate the principle of relevance, it would imply that nonscientific information and conversation did not have to be relevant and true.

Science is merely one of at least four typical modes of thought and verbal communication. While the principles of truth and relevance are undoubtedly of great importance in science and --as we have seen-- in philosophy, the role of these principles, or of certain interpretations thereof, may vary considerably in other fields of thought. In the next chapter the focus of our attention will be what characterizes the diverse modes of thought, not only philosophy and science but also literary art (as well as art in general) and especially ideology.